What TRACE Strength is
TRACE Strength is a 0–100 reading of how structurally strong a Motion market's current on-chain activity is, relative to comparable Motion tokens. It is a description of structure, not a forecast of price.
Motion answers what is moving. Trending, Top MCap and Latest are all ranked by raw magnitude — and raw magnitude is deceptive on a launchpad where a token can be nine minutes old. $500,000 of volume is extraordinary against a $50,000 market and unremarkable against a $20,000,000 one. TRACE exists to answer the next question: how strong is that movement actually?
The score is not a safety rating. A market can be structurally strong and still be exited by the people who built it. That is why Risk is computed and displayed separately, and never folded into Strength.
The cohort model
Almost every mistake in on-chain scoring comes from hardcoded thresholds. "Liquidity above $50K is healthy" is meaningless without knowing what the token is worth and how old it is. TRACE therefore scores nothing in absolute terms. Every sub-metric is converted to a percentile rank within a peer cohort.
How a cohort is chosen
A token's cohort is the set of tokens matching it on stage — still on the curve, or past the 125 HYPE target. Age and FDV bands are layered on top once the indexer supplies enough tokens per bucket.
Stage is the axis that matters most, and it is not optional. A token still on the curve holds its entire supply inside the pool, so its liquidity-to-FDV ratio is close to 100% by construction. Ranking a bonded token against that number is meaningless — it would score every mature market at the bottom of the liquidity component regardless of how deep its pool actually is. TRACE therefore ranks curve tokens only against other curve tokens, and bonded tokens only against other bonded tokens.
If fewer than 30 tokens match, TRACE widens the FDV band first, then the age band, and shows the realised cohort size on every score. A score computed against n=41 is labelled differently from one computed against n=312.
Why this matters
It means the model has no opinions to go stale. When the whole Motion market cools, percentiles recentre automatically and a 70 still means "better than roughly 70% of comparable tokens right now". The scanner learns the current shape of Motion without anyone retraining anything, and without the word "AI" appearing once.
The formula
When a component cannot be computed — Resilience needs at least three qualifying sell events — its weight is redistributed proportionally across the remaining components and the component is shown as N/A rather than as a zero. Scoring a missing measurement as zero is the difference between a model and a slot machine.
The six components
Liquidity — 20%
Not the headline liquidity number, which is trivially gamed. TRACE measures executable depth: how much HYPE can move through the pool before price moves 2% and 5%, computed by walking the V3 tick map. For a pre-bond Motion token the entire position is known — one locked range, one liquidity value — so depth is exact rather than estimated.
- 2% and 5% depth, both sides, in HYPE
- Depth ÷ FDV ratio, percentile-ranked
- Depth trend over 5 m / 1 h / 6 h
- Post-bond only: organic LP added beyond the locked launch position
Participation — 20%
Counting transactions rewards one wallet trading with itself two hundred times. TRACE counts who is participating and how evenly.
- Unique buyers per hour, cohort-normalised
- Buy-volume concentration — a Herfindahl index across wallets, inverted. One wallet at 71% of volume and two hundred wallets at 0.5% each produce the same volume and very different scores
- Holder growth net of churn
- Repeat-buyer ratio: wallets that bought more than once and have not sold
Flow — 20%
Direction and, more importantly, persistence of direction.
- Net HYPE flow over 5 m / 1 h / 6 h, exponentially weighted toward the present
- Buy/sell split measured by both count and value — they disagree more often than you would expect, and the disagreement is the signal
- Flow persistence: the share of five-minute buckets in the last hour with positive net flow. Twelve consecutive small green buckets score higher than one large one
- Damped by top-buyer contribution, so a single whale cannot manufacture a 95
Distribution — 15%
Motion tokens have a fixed 1,000,000,000 supply that begins entirely inside the locked pool position. That gives TRACE something most chains do not have: an exact figure for how much supply has genuinely left the curve. We call it absorption, and all concentration ratios are computed against it rather than against total supply.
- Top 1 / 5 / 10 / 25 share of absorbed supply
- Creator balance and creator realised sells
- Gini coefficient across holders
- Excluded from every calculation: the pool, the locked position, the Motion gateway, the zero address and known burn addresses
Motion enforces a 3% maximum wallet for the first 3,600 seconds. A wallet above 3% is therefore evidence of deliberate post-restriction accumulation, and TRACE times that against the cap-lift moment rather than treating it as generic concentration.
Momentum — 15%
The obvious mistake is scoring price +200% = 100, which turns the whole product into a pump detector. TRACE scores momentum on consistency and breadth, then classifies its shape.
- Volume acceleration: 15 m vs prior 15 m, 1 h vs prior 1 h
- Participation acceleration — is the buyer count growing, or is the same crowd trading harder?
- Trend consistency: the fraction of five-minute buckets that closed up, not the size of the move
| Classification | Condition |
|---|---|
| Building ↑ | Participation rising ahead of price |
| Accelerating ↑↑ | Volume and buyers both accelerating |
| Sustained ↑ | Consistent up-buckets, breadth holding |
| Stable → | Activity flat, no directional edge |
| Cooling ↓ | Volume decaying, price holding |
| Reversing ↓↓ | Flow negative and accelerating |
| Spike ⚡ | Extreme move on narrow participation — capped |
A Spike is explicitly capped below a Sustained move of the same magnitude. Narrow participation is a weaker structure, whatever the candle looks like.
Resilience — 10%
The most unusual component, and the one hardest to fake. TRACE asks: when people sell, does the market absorb it?
- For every sell at or above the cohort's 90th-percentile size in the last 6 h, measure price at execution and 15 minutes later — the recovery ratio
- Depth retained through those events
- Holder count through drawdown: do holders sell into weakness or hold it?
Requires at least three qualifying events. A token that has only gone up because nobody has sold yet has not demonstrated resilience — it has demonstrated nothing, and TRACE reports N/A, which is the honest answer.
Penalties
Penalties are subtracted from the weighted sum, capped at −25 in total, and always itemised. They describe observed behaviour patterns, never intent.
| Pattern | Trigger | Max |
|---|---|---|
| Concentrated flow | Top 3 buyers exceed 40% of 1 h buy volume | −6 |
| Fresh wallet cluster | Cluster of wallets first funded from a shared source within a short window | −8 |
| Round-trip activity | Repeated buy/sell cycles among the same wallet set | −10 |
| Creator distribution | Creator wallet moving or selling holdings | −8 |
| Depth decline | Executable depth down more than 25% in 1 h (post-bond) | −6 |
| Volume anomaly | Volume step-change far outside the token's own recent distribution | −5 |
Confidence
A token three minutes old and a token twelve hours old cannot support the same claim. Confidence is a function of elapsed history, trade count, unique wallet count and the number of complete five-minute buckets available.
| Level | Roughly | Display |
|---|---|---|
| Low ●○○ | Under ~20 min, or under 40 trades / 25 wallets | Score shown hatched and marked provisional |
| Medium ●●○ | Usable activity sample, limited history | Normal, with confidence marker |
| High ●●● | Enough history and breadth for full scoring | Normal |
Small-sample shrinkage
Confidence does more than annotate the score — it changes it. When the sample is thin, the honest estimate is not the raw measurement but the raw measurement pulled toward the cohort prior of 50:
A nine-minute-old market with four trades cannot honestly earn a 94, and it cannot honestly earn a 6 either. Shrinkage means TRACE reports what the data can actually support, and the token page always shows the raw figure alongside the shrunk one so nothing is hidden.
Risk, kept separate
Risk is scored 0–100 on its own axis and displayed alongside Strength, never inside it. A market can legitimately read STRENGTH 84 · RISK ELEVATED, and that combination is more informative than either number alone — it is precisely the state where a strong-looking chart is held up by a narrow base.
TRACE reports observable behaviour. It does not print "SCAM DETECTED". A flag says three wallets generated 54% of the last hour's buy volume, because that is a fact; it does not say what those wallets intend, because that is not knowable from a block explorer.
Divergence
The single most useful thing TRACE renders is the Strength line drawn on the same axis as price. Two shapes matter:
- Structure leading price — Strength climbs while price is flat. Participation, depth and flow are building before the move. This is what the Emerging tab is built to surface.
- Price leading structure — price climbs while Strength is flat or falling. The move is not supported by broadening participation or deepening liquidity. TRACE labels this an unsupported move.
Neither is a prediction. Both are descriptions of what the chain currently shows, and both are visible at a glance rather than requiring you to read six panels.
Lifecycle awareness
Motion's mechanics are unusual and TRACE is built specifically around them rather than being a generic screener pointed at a new chain.
Stage 1 — Gated
At creation, Motion deploys the token, opens a HyperSwap V3 pool at the 0.01% fee tier against WHYPE, and mints a single one-sided position holding the entire supply. That position is the bonding curve. Trades are gated through the Motion swap gateway until 125 HYPE has entered the pool, and a 3% maximum wallet is enforced for the first hour.
In this stage TRACE weights buyer breadth, HYPE inflow, wallet diversity, absorption and curve progress. Depth is closed-form from the known position — no estimation required.
Stage 2 — Bonded
After the target is reached the gate opens and the market trades directly on HyperSwap V3. Organic liquidity providers can now add and remove. TRACE shifts weight toward executable depth, depth stability, slippage, buy/sell flow and resilience — the things that only become meaningful once liquidity can leave.
What TRACE cannot see
Stating this plainly is part of the product, not a disclaimer bolted to the footer.
- Intent. TRACE sees wallets and amounts. It cannot see who controls a wallet or what they plan to do.
- Off-chain coordination. A group organising in a private channel is invisible until it touches the chain.
- Sophisticated sybils. Fresh-wallet clustering catches the cheap version. Well-funded, well-aged, independently-sourced wallets are indistinguishable from organic demand.
- Contract risk beyond Motion's template. Motion tokens are all deployed from the same audited-by-inspection template, which removes most of this — but the locker and gateway contracts are unverified source at time of writing, and TRACE says so rather than assuming.
- The future. Every number on this site describes the present tense.
Data pipeline
TRACE indexes HyperEVM (chain 999) directly. The public RPC caps eth_getLogs at 50 blocks and offers no historical state, so backfill runs from the raw block archive and a dedicated archival provider, with the public endpoint used only for tip-following.
Every token on this site is real, with its real contract address, price, liquidity, FDV and trade counts, captured from live HyperEVM market data on 25 Aug 2026. Nothing is generated. What is not yet real is the completeness of the score: Distribution and Resilience require swap-level and holder-level history, so they report N/A rather than an estimate.